MicroRNAs, or miRNAs, are small RNA molecules that help control how genes are expressed. Rather than coding for proteins themselves, miRNAs can bind to messenger RNAs, or mRNAs, and reduce the amount of protein those messages produce. Because a single miRNA can influence many different genes, changes in miRNA activity can have widespread effects on cells.

Researchers have linked altered miRNA expression to cancer, neurodegenerative disorders, and many other diseases. High-throughput miRNA sequencing, sometimes called miRNome sequencing, allows scientists to measure large numbers of miRNAs at the same time. The challenge is turning the resulting sequencing data into meaningful biological information.

A research team from Merck & Co has developed MiRQuery, an interactive web application designed to make miRNA sequencing analysis more accessible to researchers with different levels of bioinformatics experience. Yang is associated with Merck & Co., and the MiRQuery software is maintained through an MSD research GitHub repository.

Workflow of the MiRQuery application

This workflow diagram illustrates the standard analysis pipeline within MiRQuery, detailing the modules available to users and their dependencies, including exploratory visualization, differential miRNA expression analysis, target gene retrieval, pathway enrichment analysis, and correlation analysis between differentially expressed miRNAs and genes. Dotted lines indicate optional inputs or modules. Abbreviations: I- inputs, M- modules

Why microRNA sequencing requires specialized analysis

RNA sequencing can generate large amounts of information about which RNA molecules are present in a biological sample and how abundant they are. When researchers focus specifically on miRNAs, the resulting dataset can reveal patterns of small RNA expression associated with different diseases, treatments, tissues, or experimental conditions.

Generating the sequencing data, however, is only the first step. Researchers must compare samples, identify miRNAs whose expression differs between groups, visualize those differences, and determine which genes may be regulated by the miRNAs they identify.

These tasks typically require multiple bioinformatics tools. For researchers without extensive programming experience, moving between software packages and analysis environments can make the process difficult.

MiRQuery was developed to bring several of these analytical steps together in a single interactive interface.

Exploring miRNA sequencing data visually

MiRQuery is built using R Shiny, a framework that allows statistical analyses written in R to be accessed through an interactive web interface.

The application includes several common approaches for exploring sequencing data. Researchers can generate multidimensional scaling plots to examine relationships between samples, stacked column charts to compare miRNA composition, heatmaps to visualize expression patterns, and boxplots to examine individual miRNAs.

These visualizations can help researchers determine whether biological groups separate from one another, identify unusual samples, and recognize miRNAs that may deserve further investigation.

Making these tools available through a graphical interface can be particularly helpful for scientists who understand the biological questions behind their experiments but have limited experience writing analysis code.

Finding differentially expressed microRNAs

One of the central goals of miRNA sequencing is often to identify miRNAs whose abundance changes between experimental groups.

For example, researchers might compare tumor tissue with healthy tissue or compare cells before and after treatment. A miRNA that consistently increases or decreases under a particular condition could provide clues about the biological processes involved.

MiRQuery supports differential miRNA expression analysis, allowing researchers to identify these changes directly within the application.

This is important because miRNAs do not function in isolation. Their biological importance depends largely on the messenger RNAs they regulate.

Connecting miRNAs with their potential gene targets

A particularly useful feature of MiRQuery is the ability to move from differentially expressed miRNAs to their predicted gene targets.

Because one miRNA can regulate many mRNAs, an altered miRNA may influence an entire network of genes. Conversely, the same gene may be regulated by several different miRNAs.

MiRQuery allows users to retrieve predicted target genes for miRNAs identified in their analysis. Researchers can then perform pathway overrepresentation analysis to determine whether those target genes are concentrated in particular biological pathways.

For example, a group of altered miRNAs might collectively target genes involved in immune signaling, cell growth, metabolism, or programmed cell death. Examining these pathways can provide more biological context than looking at individual miRNAs alone.

Combining microRNA and messenger RNA sequencing

MiRQuery also supports integration of miRNA sequencing with bulk mRNA sequencing data.

This combination can be especially informative because miRNAs generally reduce expression of their target messenger RNAs. Researchers may therefore expect an increase in a particular miRNA to be associated with a decrease in expression of one or more target genes.

If users provide paired mRNA sequencing data, MiRQuery can identify differentially expressed genes and search for negatively correlated miRNA and gene pairs.

This allows researchers to move beyond predicting which genes a miRNA might regulate and examine whether the sequencing data show the type of relationship expected from miRNA-mediated regulation.

For instance, if a miRNA becomes more abundant while a predicted target gene becomes less abundant across the same samples, that relationship may provide additional evidence that the miRNA is influencing the gene.

Such findings would still require further experimental validation, but integrated RNA sequencing analysis can help researchers prioritize the most promising regulatory relationships for follow-up experiments.

Making bioinformatics tools easier to use

Many sequencing analysis tools are developed primarily for researchers who are comfortable working with programming languages and command-line software. That can create a barrier for molecular biologists and other laboratory scientists who generate sequencing data but do not specialize in bioinformatics.

MiRQuery attempts to reduce that barrier by placing sophisticated analytical methods behind an interactive interface.

The goal is not only to make analysis easier for scientists who are new to bioinformatics. The application can also help bioinformaticians who are new to miRNA biology by providing an organized workflow tailored specifically to miRNA sequencing.

Because the analyses are incorporated into a defined application, researchers may also be able to perform common analyses more consistently and reproducibly.

Moving from sequencing data to biological questions

The value of RNA sequencing comes from more than simply identifying which RNA molecules are present. Researchers ultimately want to understand how changes in RNA relate to biological processes and disease.

For miRNAs, that requires connecting changes in small RNA expression with the genes and pathways those molecules may regulate.

MiRQuery brings several parts of that process together, including data visualization, differential expression analysis, target prediction, pathway analysis, and integration with mRNA sequencing.

By making these tools available through a user-friendly interface, the application provides researchers with a more direct route from miRNA sequencing data to biological hypotheses that can be explored in future experiments.

Availability – MiRQuery can be accessed through PositConnect at https://julianneyang-mirquery.share.connect.posit.cloud/, and alternatively is available by user local installation via instructions on the Github project homepage (https://github.com/MSDLLCpapers/miRQuery).